Bulk superconductivity in ${\mathrm{FeTe}}_{1-x}{\mathrm{Se}}_x$ via physicochemical pumping of excess iron

The iron-based superconductor ${\mathrm{FeTe}}_{1-x}{\mathrm{Se}}_x$ has attracted considerable attention as a candidate topological superconductor owing to a unique combination of topological surface states and bulk high-temperature superconductivity. The superconducting properties of as-grown single crystals, however, are highly variable and synthesis dependent due to excess interstitial iron impurities incorporated during growth. Here we report a physicochemical process for pumping this interstitial iron out of the ${\mathrm{FeTe}}_{1-x}{\mathrm{Se}}_x$ matrix and achieving bulk superconductivity. Our method should have significant value for the synthesis of high-quality single crystals of ${\mathrm{FeTe}}_{1-x}{\mathrm{Se}}_x$ with large superconducting volume fractions.

Figure 1

XPS measurements of an air-annealed sample of ${\mathrm{Fe}}_{1+y}{\mathrm{Te}}_{1-x}{\mathrm{Se}}_x$. (a) Survey spectrum showing the presence of iron, oxygen, and carbon on the surface. Analysis of O $1s$ and Fe $2p$ peaks gives an initial Fe:O atomic ratio of 1:2.05. (b) High-resolution scan of the O $1s$ peak, from which it is determined that only 73.5% of oxygen is bonded to iron, with the remainder coming from various carbon compounds on the surface. This yields a corrected Fe:O ratio of 1:1.$51\left({\mathrm{Fe}}_2{\mathrm{O}}_3\right)$.

Figure 2

Magnetic susceptibility measured during a single post-annealing cycle for samples with thickness $\sim 75\mu \mathrm{m}$. (a) As-grown samples do not exhibit signs of superconductivity, but after annealing in air, a large diamagnetic signal is measured. The diamagnetism is suppressed by subsequent vacuum annealing. (b) The apparent superconducting volume fraction of an air-annealed sample is significantly reduced after pulverizing.

Figure 3

Evolution of magnetic susceptibility after multiple annealing cycles for samples with thickness $\sim 75\mu \mathrm{m}$. (a) Annealing in air consistently yields apparent superconducting volume fractions above 90%. (b) A steady cycle-to-cycle improvement in volume fraction is observed after the vacuum annealing steps. (c) After the final vacuum annealing step, pulverization of a single crystal no longer reduces the apparent superconducting volume fraction, indicating bulk superconductivity has been achieved.

Figure 4

Heat capacity versus temperature for an as-grown (nonsuperconducting) crystal, a crystal that has undergone a single post-annealing step and exhibits a large diamagnetic transition, and a fully cycled crystal. Annealing once results in a small change in heat capacity at the critical temperature, likely resulting from surface superconductivity, but only fully cycled crystals show a large heat capacity anomaly indicative of bulk superconductivity.

Figure 5

Number of post-annealing cycles required to reach bulk superconductivity versus crystal thickness. An approximately linear dependence is observed with a proportionality constant of $\sim 54$ cycles per millimeter of thickness.

Figure 6

STM data from a cleaved surface of ${\mathrm{FeTe}}_{0.59}{\mathrm{Se}}_{0.41}$ measured at 4.5 K. (a) Representative topographic image showing tellurium atoms (bright spots) and selenium atoms (dark spots) randomly distributed on a square lattice. Counting the spots yields a Te:Se ratio of 0.61:0.39. Also apparent are brighter spots representing interstitial iron impurities. (b) $dI/dV$ spectra taken along a cut extending away from the interstitial iron impurity marked by a red cross in (a), showing a superconducting energy gap of $\sim 2.5$ meV. STM setup conditions: $I_{\mathrm{set}}=100$ pA; $V_{\mathrm{sample}}=-10$ mV; $V_{\mathrm{exc}}=0.2$ mV (zero-to-peak).

Figure 7

Illustration of the cyclical physicochemical pumping procedure. Orange dots represent excess iron impurities. Due to the self-limiting nature of the oxidation step, after each cycle a fixed amount of excess iron is removed from the crystal. The process may be considered complete when the vacuum annealing step no longer changes the volume fraction determined from magnetic susceptibility measurements.